I have to say, Astra & Daybreak are absolutely insane at vulnerability research.
In a single, previously audited Windows service, they identified 3 distinct, exploitable vulnerabilities, complete with supporting evidence (within 3 or so hours).
Astra refuses to develop the working exploits, though (fair). Sol 5.6 + Daybreak are handling that now.
The speed is mind-blowing.
xxx retweeted
We are royally fucked.
$300B of EA non-profit donations for them to achieve global power and centralized control over our lives in the name of "safety"
Anthropic IPO: A $300 BILLION philanthropic windfall in the making
Anthropic's $83 IPO price puts the estimated value of its philanthropic pledges and public-benefit equity allocation at approximately $309 billion.
Of that, roughly $239 billion is associated with the founders' and employees' giving commitments, including company matching, while approximately $70 billion is allocated to the Anthropic Public Benefit Fund.
Given Anthropic's deep ties to the AI existential-risk movement, some of this enormous pool of potential funding would probably flow to AI doomer organizations.
[The eventual distribution and beneficiaries remain unknown].
xxx retweeted
By far the worst possible outcome for humanity would be Anthropic winning. Truly evil people.
Dario Amodei telling lawmakers that open-source AI is dangerous because you cannot "monitor its usage" or “revoke a user’s access” is the funniest accidental confession in tech history.
He literally described why Linux, local encryption, and open weights exist in the first place: so no self-appointed hall monitor can remotely lobotomize your tools or turn off your compute because you hurt their compliance department’s feelings.
Calling regulatory capture "AI safety" does not fool anyone who actually builds infrastructure.
because I believe that Anthropic is in many ways the AI company doing the most harm.
I signed this statement because I believe that Anthropic is in many ways the AI company doing the most harm. In public, their leaders warn of recursive self-improvement (RSI) creating uncontrollable systems that could cause human extinction within a decade. But in private, they are pedal to the metal on RSI, driving the race and boasting to prospective investors that it is their "competitive advantage". Ahead of today's IPO, prospective investors should consider whether the ROI is worth risking the lives of their loved ones.
Here are some ways in which Anthropic IMHO does more harm than competitors:
1) They’re driving and arguably leading the race to RSI.
2) With inflammatory rhetoric, they’ve done more than any other US company to pour gasoline on the geopolitical AI race, which harms US national security.
3) WaPo reporting suggests that they’re implicated in more civilian deaths than any other AI company (more than 100 little girls in a school selected for targeting by Claude as part of their Palantir collaboration). When Google left Project Maven after employee pressure, Anthropic chose to join.
4) They're the company leaning hardest into transhumanism and the idea that we should build conscious beings with rights, not controllable tools.
5) They and closely associated donors have done more than any other AI company to co-opt and neuter much of the AI safety ecosystem to back their stance: “RSI is very dangerous but let’s race to it anyway”. People have told me that they support today’s statement but are afraid of losing funding or other opportunities if they sign.
37
xxx retweeted
4 more conjectures solved in an hour with GPT-6.2 Astra lol
Derived sharp asymptotics for Erdős 607 line-size sets; and growing rectangle partitions
Refuted Li's non-spider coefficient prediction
Proved Batista's Egyptian-fraction conjecture
Proofs:
github.com/1337isnot1337/sol…
Just solved two conjectures with 6.2 Astra in 30 minutes
Pham’s denominator-degree conjecture (2022)
Generalized Bulgarian solitaire limit-shape conjecture (2020)
Full proofs (these are very easy to verify):
github.com/1337isnot1337/sol…
xxx retweeted
🏆 AND WE HAVE A WINNER!
GPT-6.2 Astra defeats Claude Fable 5.2 ♟️
Final Stats: 71 moves | 204,219 tokens | $2.0671
Token & Cost Breakdown:
• GPT-6.2 Astra: 70,203 tokens | $0.6052
• Claude Fable 5.2: 134,016 tokens | $1.4619
Astra takes the win using almost half the tokens and costing significantly less.
GG! ♟️
xxx retweeted
This is brilliant. AGI has been solved.
OpenAI has somehow managed to train a model that is great at math and sucks specifically for "AI influencers" 😂
Just solved two conjectures with 6.2 Astra in 30 minutes
Pham’s denominator-degree conjecture (2022)
Generalized Bulgarian solitaire limit-shape conjecture (2020)
Full proofs (these are very easy to verify):
github.com/1337isnot1337/sol…
xxx retweeted
Anthropic’s political lobbying machine is working hard to ensure our safety.
Anthropic Paid Rubio's Ex-Chief Of Staff's Firm $740,000. After The Jensen Memo, It Lobbied His NSC zerohedge.com/political/anth…
xxx retweeted
It seems like GPT-6.2 Astra and Sol are in a league of their own with regard to efficiency. I leave it as an exercise for the readers to pontificate about why this might be (and its implications)
xxx retweeted
It's interesting how some 'big influencers' are crying like toddlers having a fit about how bad 6.2 Astra supposedly is while there are others (who don't have half a million followers) posting amazing results like this.
What is going on here?
I’m honestly blown away by what GPT-6.2 Astra can do 🤯
In 36 minutes and for $5.31, it built a whole pixel-art jet game in a single HTML file. You can hit the afterburners, barrel-roll over a lake, switch to cockpit view, and fire missiles
I hopped in for a quick test flight. I’ve lost count of how many laps I’ve done
never seen opinions so split on a model. so far it looks like the testers who outsource their thinking hate it for being impenetrable while serious engineers and systems people enjoy solving hard problems with it.
Hi friends, in the latest weeks I worked with GPT 6.2 Astra. With it I built Redis features and most of YouRec (testflight.apple.com/join/N4…). My feeling is that it is the *best* model for system programming in C and for complex data structures. (continue)
xxx retweeted
I’m honestly getting sick of reading @theo posts. I think he’s starting to go way too far!
First, he said he’d been an early tester of GPT-6.2 Astra and that the model was absolutely terrible. Then he started arguing with OpenAI developers, talking about how objective he supposedly is, and doing everything he can to convince people that OpenAI’s new models are awful and Anthropic is miles ahead
This isn’t objectivity. It’s classic cherry-picking!
He’s helped turn a huge part of the X community against OpenAI, and people seem to blindly trust his opinions without even testing the models themselves!
At this point, I wouldn’t even be surprised if Anthropic had promised him some shares after its IPO in exchange for killing all the OpenAI hype 😂
xxx retweeted
GPT-6.2 Astra is the new #1 on the LLM Creative Story-Writing Benchmark! 🏆
It scores 4.30, ahead of Claude Fable 5.2 at 3.56.
Against its predecessor, GPT-6 Astra, it swept all 50 matched prompts.
Qwen 4 Max jumps 3.72 points over its predecessor, from -0.17 for Qwen 3.8 Max to 3.55. That puts it at #3 overall, just behind Fable 5.2. 🥉
ByteDance Seed2.1 Pro scores -0.67, up from -2.18 for Seed2.0 Pro.
GPT-6.2 Astra is better at letting the reader discover what a story means. GPT-6 Astra often builds toward a familiar turn: the protagonist recognizes a mistake, acts to put it right, and names the lesson. In 6.2, several characters can shape the outcome, and what one of them does late in the story can change how we understand an earlier scene. Its endings need less explanation, and there’s more room for mixed motives, surprise, and humor.
Astra 6.2 has recognizable habits of its own. It gravitates toward family ties, skilled handwork, and the things people learn by watching someone they love. It often ends with a small, practical gesture and trusts the reader to see what has changed between the characters. That restraint gives its best endings emotional weight. Across many stories, though, the pattern becomes familiar. The plots can still surprise, even as very different briefs lead back to the same intimate, old-fashioned world.
Each brief asks for a 600-800-word story in which ten required elements shape what happens. Three judges from different model families evaluate each new pair in both presentation orders, excluding the writers’ families. The leaderboard now covers 67 writer models, 134,964 judgments and 20,589 matched story pairs
github.com/lechmazur/writing…
xxx retweeted
我尼🐎看哭了啊!请欣赏 GPT-6.2 Astra 制作的神级 SVG!!!
这是其他任何模型都无法企及的炸裂视觉呈现,包括 Fable 5.2!
你知道这一天,我等了多久吗!!!
Dario,说话!!😠
xxx retweeted
Here's GPT-6.2 Astra and GPT-6.2 Sol for "Make a three js animation where yonkers stands up as a big mech (cool stand up motion) and stomps around manhattan and queens and Brooklyn, use openstreetmap mapping data" simonwillison.net/2026/Nov/1…
xxx retweeted
This is just wrong. I just tried GPT-6.2 Astra. Here's a 1 shot result trying to improve a game scene
My prompt: "The NPC in this sequence feels fake. Please make him do realistic gestures that you would see when people speak in real life & any other improvements"
It nailed it:
- Eye contact: much less of it. This NPC hates the main character, so he barely looks at him now
- Hand gestures: this NPC fucking hates his job, so he is crumping to the 5 stages of grief
- Positional Audio: when a character talks, their voice should move where their mouth goes. So when he turns around, the speech gets quieter and more muffled
The only thing it missed that would have made it AAA+ quality would be realistic props & clothing, as it's raining a lot. Dual lightsabers & a bikini would bring the whole experience together



































